Aeolian Dust Dispersal Patterns Since the Last Glacial Period in Eastern

1 Aeolian dust dispersal patterns since the last glacial period in eastern

2 Central Asia: Insights from a loess-paleosol sequence in the Ili Basin

3 Yue Li1,2,3, Yougui Song1*, Kathryn E. Fitzsimmons2, Hong Chang1, Rustam Orozbaev4,5, Xinxin 4 Li 1,2 5 1 State Key Laboratory of Loess and Quaternary Geology, Institute of Earth Environment, Chinese 6 Academy of Sciences, Xi’an, 710061, China 7 2 Research Group for Terrestrial Palaeoclimates, Max Planck Institute for Chemistry, Mainz, 55128, 8 Germany 9 3 College of Earth Science, University of Chinese Academy of Sciences, Beijing, 100049, China 10 4 Research Center for Ecology and Environment of Central Asia, Chinese Academy of Sciences, 11 Urumqi, 830011, China 12 5 Institute of Geology, National Academy of Sciences, Bishkek, 720040, Kyrgyzstan 13 14 * Corresponding author: 15 Dr. Yougui Song 16 E-mail: [email protected] 17 State Key Laboratory of Loess and Quaternary Geology 18 Institute of Earth Environment, Chinese Academy of Sciences 19 No. 97 Yanxiang Road, Yanta, Xi’an 710061, China 20 Tel. 86-29 6233 6216 21 Fax. 86-29 6233 6216 22

23 Abstract 24 The extensive loess deposits of the Eurasian mid-latitudes provide important terrestrial 25 archives of Quaternary climatic change. As yet, however, loess records in Central Asia are poorly 26 understood. Here we investigate the grain size and magnetic characteristics of loess from the Nilka 27 (NLK) section in the Ili Basin of eastern Central Asia. Weak pedogenesis suggested by frequency-

28 dependent magnetic susceptibility (χfd%) and magnetic susceptibility (MS) peaks in primary loess 29 suggest that MS is more strongly influenced by allogenetic magnetic minerals than pedogenesis, 30 and may therefore be used to indicate wind strength. This is supported by the close correlation 31 between variations in MS and proportions of the sand-sized fraction. To further explore the temporal 32 variability in dust transport patterns, we identified three grain size end members (EM1, mode size 33 47.5 µm; EM2, 33.6 µm; EM3, 18.9 µm ) which represent distinct aerodynamic environments. EM1 34 and EM2 are inferred to represent grain-size fractions transported from proximal sources in short- 35 term, near-surface suspension during dust outbreaks. EM3 appears to represent a continuous 36 background dust fraction under non-dust storm conditions. Of the three end members, EM1 is most 37 likely the most sensitive recorder of wind strength. We compare our EM1 proportions and mean 38 grain size from NLK with the Jingyuan section in the Chinese loess plateau, and assess these in the 39 context of modern and Holocene climate data, and suggest that the Siberian high-pressure system is 40 the dominant influence on wind dynamics and thus loess deposition in the eastern Ili Basin. Six 41 millennial-scale cooling (Heinrich) events can be identified in the NLK loess records. Our grain- 42 size data support the hypothesis that the Siberian High acts as teleconnection between the climatic 43 systems of the North Atlantic and East Asia in the high northern latitudes, but not for the mid- 44 latitude westerlies. 45 46 Key words: Last glacial, Ili Basin, Central Asia, loess, magnetic susceptibility, grain size, 47 paleoclimate 48 49 1 Introduction 50 Central Eurasia experiences extremely continental climatic conditions, in large part due to its 51 position far from the oceans. Arid Central Asia (ACA), the mid-latitude region spanning the Caspian 52 Sea across to the eastern Tien Shan mountains, is therefore a sensitive recorder of past climate 53 change due to its location in the transitional region between the Asian monsoons (Dettman et al., 54 2001;Cheng et al., 2012), mid-latitude westerlies (Vandenberghe et al., 2006) and North Asian polar 55 front (Machalett et al., 2008). The relative importance and intensity of these major climate 56 subsystems has varied across the latitudinal and longitudinal range of Central Asia through time. 57 Thus identification of the predominant climate regimes in this region, using geological archives, is 58 a crucial precondition for tracing paleoclimatic evolution. 59 One of the most promising potential palaeoenvironmental archives in ACA is its widespread, 60 thick loess deposits. Loess is one of the most important archives of Quaternary climate change 61 (Maher, 2016;Muhs, 2013). In Central Asia, the loess deposits drape the piedmont slopes of the 62 major mountain ranges - the Tian Shan, Alai, Altai and Pamirs - from Xinjiang province in China 63 through Kazakhstan, Kyrgyzstan and Uzbekistan, into Tajikistan. While recent years have witnessed 64 increasing loess-based datasets in the region (Dodonov et al., 2006;Feng et al., 2011;Li et al., 65 2016c;Li et al., 2016b;Machalett et al., 2006;Smalley et al., 2006;Song et al., 2014;Song et al., 66 2015;Song et al., 2012;Yang et al., 2006;Youn et al., 2014;Fitzsimmons et al., 2016), the forcing

67 mechanisms and the climatic conditions responsible for loess-paleosol sequences formation are as 68 yet not systematically understood (Fitzsimmons et al., 2016;Machalett et al., 2008;Li et al., 69 2016a;Song et al., 2017b). 70 Evidence for millennial-scale climatic oscillations associated with the Greenland 71 (Dansgaard/Oeschger (D-O) events) (Dansgaard et al., 1993) and cool phases associated with 72 iceberg calving into the North Atlantic (Heinrich (H) events) (Bond et al., 1992) have been found 73 in the form of grain-size variations in loess deposits in Chinese Loess Plateau (CLP) (Sun et al., 74 2012;Porter and An, 1995;Chen et al., 1997b) and Europe (Antoine et al., 2009;Rousseau et al., 75 2007;Zeeden et al., 2016). Data for Central Asian loess do not as yet exist at this resolution (Li et 76 al., 2016b;Song et al., 2017a), despite its strategic location as a likely environmental bridge between 77 the North Atlantic and East Asian Monsoon climatic regions. 78 The Ili Basin of Central Asia hosts thick loess deposits in the strategic central eastern part of 79 ACA (Song et al., 2014). The basin is surrounded to the south and north by the Tian Shan mountain 80 range, widens to the west and drains into endorheic Lake Balkhash (Fig. 1), and provides a 81 conducive situation for loess accumulation. In this paper we present new data on the physical 82 properties of a 20.4 m thick loess deposit at Nilka (NLK) in the eastern Ili Basin. We investigate 83 variations in grain size distributions and magnetic properties in order to investigate likely links with 84 environmental dynamics. 85 2 Physical geography 86 The Ili Basin (78° ~ 85° E and 42° 30ʹ ~ 44° 30ʹ N) straddles southeast Kazakhstan and 87 northwest China. It is an intermontane basin opening westward towards the Ili drains into Lake 88 Balkhash which is in the semi-arid transitional region between the steppe and full deserts of Central 89 Asia. The Northern and Southern Tian Shan form the northern and southern margins of the basin 90 (Fig. 1a). 91 This region has a semi-arid, continental climate, with a strong precipitation gradient dependent 92 on altitude. The altitude of the basin floor is 500 ~ 780 m; the northern Tian Shan Range reaches 93 altitudes of > 4000 m a.s.l. and the southern Tian Shan mountains range between 3000 ~7000 m 94 a.s.l. towards the catchment divide. Mean annual precipitation (MAP) ranges between 200-500 mm 95 on the plains, and mean annual temperature (MAT) ranges from 2.6-10.4°C (Li, 1991;Ye, 1999). 96 The surface vegetation in this region is dominated by Desert Steppe and Steppe and the zonal soils 97 comprise Sierozem, Castonozem and Chernozem. 98 Modern meteorological data (2009 ‒ 2013) show a MAP of 354 mm and a MAT of 7.3°C in 99 Nilka site (data from the China Meteorological Data Network: http://data.cma.cn/). 100 101 Fig. 1 The location of study area and the photo of Nilka (NLK) section. 102 103 3 Materials and methods 104 3.1 Section and sampling 105 The Nilka (NLK) section (83.25°E, 43.76°N, 1253 m a.s.l) is situated on the second terrace of 106 the right bank of the Kashi River, a tributary of the Ili River. The site is located in the eastern Ili 107 Basin of far western China, adjoining the Northern Tian Shan to the north (Fig. 1b). The section has 108 a thickness of 20.4 m and overlies fluvial sands and gravels (Fig.1c). The profile has been exposed 109 by local residents for making bricks, and recently formed the focus of a geochronological study 110 comparing luminescence with radiocarbon methods (Song et al., 2015). According to the dating

111 results, the NLK loess started to accumulate since ~ 70 ka. Stratigraphically and geochronologically, 112 the loess package at NLK is equivalent to the L1 loess unit (also known as Malan loess) and S0 113 paleosol unit (known as Holocene Heilu soil) in the CLP (Liu, 1985a). Although largely 114 homogeneous in appearance, two weak paleosols (at 5.0 – 7.0 m and 15.7 – 18.0 m depths) were 115 identified in the section by field observations. We therefore divided the NLK stratigraphy into S0, 116 L1L1, L1S1, L1L2, L1S2 and L1L3 units (Fig. 1c). 117 After cleaning the NLK section to remove dry, weathered sediment, samples were collected at 118 intervals of 2 cm. A total of 1026 bulk samples were prepared for measurements of physical 119 characteristics. Because the optically stimulated luminescence (OSL) dating is more reliable for 120 constructing a loess chronology than bulk sediment AMS 14C dates (Song et al., 2015), this study 121 uses the previously published OSL dating results as the basis for our age model. 122 3.2 Grain-size analyses 123 Prior to grain size measurements, 0.5 g of dry bulk sample was pretreated by the removal of

124 organic matter and carbonate using H2O2 and HCl, respectively (Lu and An, 1997). Samples were

125 then dispersed in an ultrasonic bath for 5 min with 10 ml 10% (NaPO3)6 solution. Grain size 126 distribution was analyzed using a Malvern 2000 laser instrument at the State Key Laboratory of 127 Loess and Quaternary Geology, Institute of Earth Environment, Chinese Academy of Sciences. 128 Particle size distribution was calculated for 100 grain size classes within a measuring range of 129 0.02−2000 μm. Replicate analyses indicated an analytical error of < 2% for the mean grain size. 130 End-member unmixing of loess grain-size distributions is based on the hierarchical Bayesian 131 model for end-member modeling analysis (BEMMA; Yu et al. (2016)). Grain-size parameters were 132 calculated from the analytical data with GRADISTAT (Version 4.0; Blott (2000)). 133 Two samples (NLK1106 at 11.06 m and NLK1840 at 17.8 m) were also selected for the 134 extraction and measurement of mineral-specific quartz grain size according to published methods 135 (Sun et al., 2000a). The isolated quartz grain samples (Fig. S1) were then analyzed by the Malvern 136 2000 laser instrument so that comparisons of quartz grain and bulk samples could be performed to 137 investigate the weathering degree of NLK loess. 138 3.3 Magnetic susceptibility measurements 139 Magnetic susceptibility was measured with a Bartington MS2 meter at the State Key laboratory 140 of Loess and Quaternary Geology, Institute of Earth Environment, Chinese Academy of Sciences. 141 Samples were oven-dried at 40°C for 24 hours. Subsamples of 10 g from each sample were then 142 precisely weighed for magnetic measurements. Low- (0.47 kHz) and high- (4.7 kHz) frequency

143 magnetic susceptibility (χlf and χhf, respectively) were measured. The absolute frequency-dependent

144 magnetic susceptibility was calculated as χfd = χlf − χhf. Frequency-dependent magnetic susceptibility

145 was defined and calculated as χfd % = [(χlf − χhf)/ χlf] ×100%. 146 4 Results 147 4.1 Magnetic susceptibility variations 148 Both magnetic susceptibility (MS) data and stratigraphy show a close correspondence 149 throughout the NLK section. We observe higher MS values within primary loess and lower values 150 within paleosols. The exception to this trend is the modern (S0) soil which yields high MS values 151 (Fig. 2). 152

153 Fig. 2 Lithology and magnetic susceptibility characteristics (χlf, χfd and χfd%) of the NLK section. 154

- 155 The χlf values of the S0 unit are higher than for the L1 unit, with an average of 98.13 × 10 8 3 -1 -8 3 -1 156 m kg . The χlf values of the L1L1 unit vary from 56.5 − 103.9 × 10 m kg , decreasing down-

157 profile. The χlf value abruptly decreases at c. 5 m, with generally lower values in the L1S1 unit, -8 3 -1 158 averaging 62.58 × 10 m kg . χlf in the L1L2 unit gradually increases down profile, with significant -8 3 -1 159 fluctuations in the lower part; χlf values vary from 67 − 102.55 × 10 m kg . Lower χlf values are -8 3 -1 160 observed in L1S1 unit with an average value of 57.99 × 10 m kg . In the L1L3 unit, the χlf values 161 vary with greater amplitude around an average value of 68.74 × 10-8m3kg-1.

162 Absolute frequency-dependent magnetic susceptibility (χfd) values likewise vary with

163 stratigraphy. The S0 unit yields the highest χfd value. The L1 unit is characterized by relatively

164 consistent and lower χfd values. Frequency-dependent magnetic susceptibility (χfd%) yields the same

165 trend as χfd, although χfd% values clearly increase in the central part of L1S2. 166 4.2 Mixing model of loess grain-size distributions 167 The mean grain-size distribution, and variation range of volume frequencies for each grain- 168 size class in the dataset, are presented in Fig. 3a. The overall grain-size frequency curve shows a 169 unimodal pattern, slightly skewed towards the coarser side, with the primary mode ranging from 170 11.9 – 47.5 µm. An additional small grain size peak occurs at c. 0.4 – 2 µm. Three unmixed end 171 members were identified (Fig. S2), yielding fine-skewed grain-size distributions with clearly 172 defined modes of 47.5 µm (EM1), 33.6 µm (EM2) and 18.9 µm (EM3) (Fig. 3b). 173 174 Fig. 3 End-member modelling results of the grain-size dataset of the NLK section. (a) Mean size 175 distribution and range of volume frequency for each size class. (b) Modelled end-members 176 according to the three-end-member model (modal sizes: ~ 47.5 µm, ~ 33.6 µm and ~ 18.9 µm). 177 Size limits of clay, silt and sand fractions determined by laser particle sizer differ from those 178 derived by the pipette method. The upper limits of grain-size classes used here are at 4.6/5.5 µm 179 for clay, 26 µm for fine silt, and 52 µm for coarse silt, as previously published by Konert and 180 Vandenberghe (1997). Sand is designated for particle sizes > 52 µm. Therefore, EM1 and EM2 181 correspond to coarse silt and EM3 to fine silt. 182 183 The proportional distribution of the end members down the section is shown in Fig. 4. In the 184 primary loess units (L1L1, L1L2 and L1L3), the deposits are dominated by the coarser silt EM1 and 185 EM2, while higher proportions of fine silt EM3 are observed within the soil horizons (S0, L1S1 and 186 L1S2). EM1 displays high frequency, large amplitude fluctuations down the profile, varying 187 between 0.09 – 0.72, and clearly dominates the primary loess units and occurs in low proportions in 188 the soil units (Fig. 4). EM2 shows a similar trend to EM1, but with less variability down profile. 189 Proportions of EM2 range between 0.11 − 0.66 with minimal fluctuations within individual units, 190 and proportions decrease significantly in the soil units S0 and L1S2. Proportions of EM3 remain 191 consistently low within the primary loess units, and increase to 0.46 and 0.8 within the soil horizons 192 S0 and L1S2 respectively. 193 194 Fig. 4 Proportional contributions of the three end-members in the NLK section. 195 196 5 Discussion 197 5.1 Impacts of wind strength on magnetic susceptibility variations 198 Magnetic susceptibility (MS) in loess is predominantly determined by the concentration of

199 iron-bearing magnetic minerals within the sediment (Liu et al., 1999;Liu et al., 1994;Song et al., 200 2010). Generally, this varies between primary loess and soil horizons. Soils generally experience an 201 enrichment in magnetic minerals (higher MS), compared with primary loess (Zhou et al., 202 1990;Maher and Thompson, 1992;Heller and Evans, 1995;Heller and Liu, 1984;Ding et al., 203 2002;Buggle et al., 2009). 204 The contrast between high and low MS in paleosols and primary loess, respectively, typically 205 forms the basis for the stratigraphic differentiation of loess deposits. The main MS variations in the 206 NLK loess sequence, with the exception of the S0 unit, however, do not follow this trend (Fig. 2). 207 At NLK, lower MS values are found in the paleosols and higher MS in loess units. A similar case 208 also occurs in the L1 loess layers at other sites in the Ili Valley, such as the TLD, ZKT and AXK 209 sections (Fig. 1) (Jia et al., 2010;Jia et al., 2012;Song et al., 2010).

210 χfd indicates the concentration of magnetic particles within a small grain size range across the 211 superparamagnetic (SP)/stable single domain (SSD) boundary (Liu et al., 2012) (magnetite, ˂ ~100 212 nm; maghemite, ˂ ~20 µm). Particles with this grain size are considered to form in situ within soils

213 during pedogenesis (Maher and Taylor, 1988;Zhou et al., 1990). Therefore, χfd can serve as a direct 214 proxy for pedogenesis (Heller et al., 1993;Maher and Thompson, 1995;Liu et al., 2007;Buggle et

215 al., 2014). In the NLK section, χfd yields consistently low values throughout the sequence and

216 indicates no clear enrichment even in the paleosol layers L1S1 and L1S2. Comparison between χlf

217 vs. χfd down profile shows no correlation between MS and SP particles (Fig, S3c). These results 218 suggest that SP particles played only a minor role in MS enhancement in the NLK loess.

219 χfd% is used as a proxy to determine the contribution of SP particles to MS (Zhou et al.,

220 1990;Liu et al., 1992). At NLK section, however, we also observe consistently low χfd% values in 221 both loess and paleosol layers, with a slight increase only in the L1S1 paleosol. This observation 222 reinforces our interpretation that the content of SP particles is very low, and consequently that their 223 contribution to MS can be ignored. 224 The low proportions of SP particles in the NLK loess imply that the pseudo-single-domain 225 (PSD) and multi-domain (MD) magnetic grains, rather than SP grains forming in situ, are more 226 influential for the magnetic enhancement at this site. Since PSD and MD magnetic minerals are 227 difficult to produce during pedogenesis (Song et al., 2010), such minerals are more likely to be 228 detrital in nature, deriving from the original protolith. 229

230 Fig. 5 Comparison of different grain size fractions of NLK loess with χlf (limits of grain-size classes 231 after Konert and Vandenberghe (1997) ). 232 233 In some cases, moist conditions associated with pedogenesis may result in the weathering and 234 dissolution of the magnetic minerals maghemite and magnetite (Nawrocki et al., 1996;Maher, 235 1998;Grimley and Arruda, 2007). In such cases, a negative relationship between magnetic

236 susceptibility and pedogenesis can develop, in contrast to the classical situation whereby χfd is 237 enhanced. At NLK section, however, we observe no textures caused by groundwater fluctuations. 238 We therefore exclude groundwater fluctuations and high levels of precipitation as a factor in our 239 MS characteristics at NLK section. 240 In the wind velocity/vigor model (also known as the Alaskan or Siberian model), wind strength 241 affects MS values of loess through the physical sorting of magnetic grains (Beget and Hawkins, 242 1989). The influence of this process on MS values in loess can be assessed by investigating the

243 correlation between MS and coarser (silt or sand) and finer clay percentages (Fig. S3). At NLK, low 244 MS values in the S0 soil between 0 – 0.5 m correlate positively with clay percentage variations (Fig. 245 S3a), while higher MS values at depths greater than > 0.5 m correlate closely with increased sand 246 concentrations (Fig. S3b). We therefore propose that MS enhancement at NLK is likely driven by 247 wind strength. Under this scenario, weak pedogenesis prevented the efficient production of SP 248 grains (Fig. 2), and allogenetic magnetic minerals associated with dust transportation made a greater 249 contribution to the MS. Wind strength can therefore be interpreted as the main influence on MS 250 variations at NLK. The enhancement of magnetic susceptibility in NLK loess most likely falls into 251 region A in Fig. 9 of Liu et al. (2013). Region A represents the area where the climate is arid, and 252 pedogenesis is weak and dominated by physical weathering (Liu et al., 2013). Therefore, we can 253 use the “wind theory” to decipher the MS variations of NLK loess. Weak pedogenesis enables 254 preservation of primary atmospheric dust contributions to NLK. 255 5.2 Genetic interpretations of end members in loess grain size 256 Grain-size analysis was conducted in order to understand wind dynamics (strength and 257 direction) during loess deposition (Liu, 1985b;Lu and An, 1998;Sun et al., 2010). Grain-size 258 analysis provides information on sediment depositional mechanisms as well as an insight into 259 spatio-temporal changes in deposition, provided factors such as vegetation, pedogenesis, grain size 260 of source sediments, and distance from the deposition area to source area are taken into account 261 (Qin et al., 2005;DiPietro et al., 2017;Obreht et al., 2015;Terhorst et al., 2012;Ding et al., 2005;Ding 262 et al., 1999;Yang and Ding, 2008). 263 Statistical analysis of loess grain-size offers new opportunities for understanding paleoclimate 264 variations. Studies increasingly use grain-size partitioning to identify sub-populations within bulk 265 samples. There are two dominant approaches to unmixing grain-size spectra: parametric 266 decomposition (e.g. Sun et al., 2002) and non-parametric decomposition (e.g. Prins and Vriend, 267 2007;Weltje, 1997; Weltje and Prins, 2007) . Based on the statistical datasets generated, the different 268 end members can be interpreted to infer distinct atmospheric transport mechanisms, modes and 269 travel distances (Ujvari et al., 2016). In some cases, the end-member approach has been used to 270 identify variation in the geological context or source area (Prins et al., 2007). We investigated the 271 applicability of this approach to the Ili Basin loess at NLK by unmixing grain-size distributions with 272 BEMMA (Yu et al., 2016), generating a mixing model consisting of three end members (Fig. S2). 273 Relying largely on samples from the Eurasian loess belt extending from the Russian Plain north 274 of the Caspian Sea eastwards to the Tibetan Plateau and CLP, Vandenberghe (2013) applied the 275 visual inspection of grain-size distribution curves and EMMA end-member analysis to define the 276 characteristic grain-size distribution of primary loess deposits and interpret the likely conditions of 277 transport and deposition. Using the sediment groups identified in Vandenberghe (2013), some 278 studies interpreted multiple sources for loess sediments (Yang et al., 2016;Nottebaum et al., 279 2015;Nottebaum et al., 2014). In this study, we apply the end-member analysis of the NLK loess to 280 the sediment groups of Vandenberghe (2013) in an effort to reconstruct dominant aeolian processes. 281 Fine sand (‘sediment type 1.a’ in Vandenberghe (2013)) is a typical component of loess deposits 282 near to or overlying river terraces. Although the NLK section lies on the second terrace of the Kashi 283 River and therefore close to a potential source of coarser grained material, the fine-sand end member 284 is completely absent. Modal grain sizes in this range (c. 75 µm ) are common in loess along the 285 Huang Shui and Yellow Rivers in China (Vriend and Prins, 2005;Vandenberghe et al., 2006;Prins et 286 al., 2009), the Danube and Tisza rivers in Serbia (Bokhorst et al., 2011), and the Mississippi valley

287 in the USA (Jacobs et al., 2011). Since this fraction is interpreted to originate from proximal sources, 288 the grain size of the available source material, rather than wind energy, plays a more important role 289 in the presence and proportions of this grain size (Vandenberghe, 2013). The lack of fine sand at 290 NLK may be attributed to 1) its location in the upper reaches of the Kashi river (Fig. 1b) in a region 291 which lacks available supplies of fine sand; 2) the V-shaped nature of the channel which is not 292 conducive to aeolian entrainment of bank deposits; and 3) the relatively high altitude of NLK within 293 the basin which inhibits transport and deposition of coarser sediment grains (Vandenberghe, 2013). 294 The three members identified at NLK correspond to coarse silt (EM1 and EM2) and fine silt 295 (EM3) (Fig. 3b). Each likely represent different kinds of depositional processes which operated 296 throughout the accumulation of the deposit at NLK. Here we focus on the implications of these three 297 end members for understanding past environmental conditions responsible for loess-paleosol 298 sequences formation. 299 EM1 has a modal grain size of 47.5 µm (Fig. 3b), which approximately corresponds to the 300 ‘subgroup 1.b.1’ of Vandenberghe (2013). The mode is similar to end members identified in loess 301 from the CLP and north-eastern Tibetan Plateau (NE-TP) (EM-2: 44 µm) (Vriend et al., 2011). The 302 size of this component is unlikely to be due to longer distance transport. Rather, it is inferred to 303 derive from shorter distance transport of suspended load (Vriend et al., 2011;Vandenberghe, 2013). 304 Coarser particles (>20 µm ) rarely reach suspension above the near surface (0 − 200 m above the 305 ground). When entrained by wind, they do not remain in suspension for long enough to travel long 306 distances (Tsoar and Pye, 1987;Pye, 1987). Since the average grain-size of EM1 is 26.74 µm 307 (calculated after Folk and Ward (1957)), we infer that this fraction was transported mainly during 308 short-term suspension episodes at lower elevations by surface winds, and deposited short distances 309 downwind of the source. These short-term suspension episodes may correspond to spring-summer 310 dust storms. Our interpretation is supported by present-day dust measurements on the CLP which 311 identify a similar modal grain-size during such events (Sun et al., 2003). 312 EM2 represents a mode at 33.6 µm (Fig. 3b). It lies towards the finer end of the range of 313 ‘subgroup 1.b.2’ (Vandenberghe, 2013). Comparable loess of the same grain size has been identified 314 in loess from the northern Qilian Shan/Hexi Corridor (EM2: 33 µm) in northern China, which was 315 also interpreted as depositing from short-term suspension (Nottebaum et al., 2015). Loess of this 316 grain size has been attributed to dust fallout (Pye, 1995;Muhs and Bettis, 2003) and fallout from 317 low-altitude suspension clouds (Sun et al., 2003), as measured from modern depositional events. 318 This fraction requires less wind energy than EM1, is transported further, is more widely distributed, 319 and therefore comprises a higher proportion of distal loess populations (Vandenberghe, 2013). We 320 propose that EM2 was transported mainly in short-term, near-surface suspension during dust storms, 321 and that wind strength controlled the relative proportions of EM1 and EM2 through time (see the 322 mirror image relationships over millennial scales in Fig. 4). This interpretation implies that both 323 EM1 and EM2 have the same origin. 324 The grain-size distribution of EM3 has a modal peak at 18.9 µm (Fig. 3b). This population 325 belongs to ‘subgroup 1.c.1’ in Vandenberghe (2013). This population is also widespread in loess 326 from the CLP and NE-TP (Prins et al., 2007;Prins and Vriend, 2007), the Danube Basin loess of 327 Europe (Bokhorst et al., 2011;Varga, 2011). It is particularly common in loess of interglacial age 328 (Vriend, 2007). There is as yet no consensus regarding the transport processes responsible for this 329 grain size population. On the one hand, researchers have suggested that grains of this size can be 330 lifted by strong vertical air movement and subsequently incorporated into the high-level westerly

331 air streams (Pye, 1995;Pye and Zhou, 1989). This process would link EM3 with long-term 332 suspension transport driven by high-level Westerlies (Prins et al., 2007;Vriend et al., 333 2011;Nottebaum et al., 2014;Vandenberghe, 2013). Conversely, Zhang et al. (1999) argued that < 334 20 µm particle fractions derives from “non-dust storm processes” associated with north-westerly 335 surface winds. We argue for the latter on the basis that the EM3 modal grain size from the CLP and 336 NE-TP is coarser (Vriend et al., 2011) than EM3 at NLK in the Ili Valley, which is located further 337 west. If EM3 was transported by high-level westerlies, then one would expect either no significant 338 change (Rea et al., 1985;Rea and Hovan, 1995), or a decrease in grain size from west to east 339 concomitant with wind direction. Furthermore, with mathematical fitting, Sun et al. (2004) related 340 a fine component (2 – 8 µm) to high-altitude westerlies. This fine component is comparable to 341 ‘subgroup 1.c.2’ of Vandenberghe (2013) but is not consistent with our modal size of EM3. 342 Observations of modern aeolian processes at the southern margins of the Tarim Basin indicate that 343 fine grain sizes similar to EM3 (8 − 15 µm) are deposited by settling during low velocity wind 344 conditions (Lin et al., 2016). Particle-size distributions of background dust from the northern slopes 345 of the Tianshan Mountains also typically yield a modal peak of approximately 10 µm (Schettler et 346 al., 2014). We therefore infer the EM3 modal peak to derive from low altitude, non-dust storm 347 processes. 348 Fine particles can also be incorporated into silt- or sand-sized aggregates which can be 349 transported by a range of wind velocities, including dust storms (Qiang et al., 2010;Pye, 350 1995;Derbyshire et al., 1998;Mason et al., 2003). For example, Ujvari et al. (2016) argued that ~ 1 351 – 20 µm fractions are affected by aggregation, as shown by comparison between minimally and 352 fully dispersed grain size distribution measurements of loess samples from southern Hungary. Under 353 higher wind velocity conditions, aggregates should co-vary with the coarser EM1 particles 354 transported by surface winds during dust storms. However, since this model is unlikely to hold for 355 EM3 particles (Fig. 4), the aggregate model is unlikely to be responsible for the presence of EM3 356 grain sizes at NLK. 357 Post-depositional processes may also influence grain-size distribution. In large part this occurs 358 due to chemical weathering which produces very fine silt and clay minerals (Xiao et al., 1995;Wang 359 et al., 2006;Hao et al., 2008). Quartz grains are more resistant to weathering and remain largely 360 unaltered during post-depositional processes. Consequently, quartz mineral grain size may be used 361 as a more reliable proxy indicator of winter monsoon strength than other components (Sun et al., 362 2006;Sun et al., 2000b;Xiao et al., 1995). Figure. 6a shows the grain-size distribution curves of 363 quartz grains isolated from primary loess (yellow) and paleosol (red) samples. The quartz modal 364 grain size is finer in the paleosol than in the primary loess unit. From this we can deduce that wind 365 strength was weaker during pedogenesis, and stronger during periods of primary loess deposition. 366 The grain size distributions of bulk samples display similar characteristics with those of quartz 367 samples mentioned above (Fig. 6b), since soil unit modal peaks (red and orange) are finer than those 368 in the primary loess (blue and green). Therefore, we argue that wind strength, rather than the post- 369 depositional pedogenesis, has the greatest influence on grain size distribution at NLK, and that EM3 370 was not produced by chemical weathering. 371 372 Fig. 6 Comparison of grain size distribution between purified quartz subsamples of paleosol and 373 primary loess (a), and between bulk samples of paleosols and primary loess (b). Comparison of 374 the grain size distribution between EM3 and samples from weak paleosol units (c).

375 376 The relative proportions of the end members down profile can yield information about temporal 377 variability in wind dynamics. The fairly consistent proportions of EM3 within the loess units 378 indicate it to represent continuous background dust through time (Vandenberghe, 2013). Proportions 379 of EM1 and EM2 decrease noticeably within paleosol units relative to EM3 (Fig. 4). This indicates 380 that variations in proportions of EM3 are mainly driven by variability in EM1 and EM2 (Vriend et 381 al., 2011), but also that consistent background sedimentation of EM3 continued during weak 382 pedogenesis (Fig. 6c). This characteristic is comparable with observations from the CLP (Zhang et 383 al., 1999). 384 In addition, small peaks at c. 0.8 µm are also observed in the grain-size distribution curves of 385 all three end members. The generation of these finest grain peaks may be due to post-depositional 386 pedogenesis (Sun, 2006), especially for particles < 2 µm (Bronger and Heinkele, 1990;Sun, 2006). 387 However, since the dominant modal peaks are much coarser, weaker post-depositional weathering 388 as suggested by MS is unlikely to have had a significant influence on the populations of EM1, EM2 389 or EM3 at NLK. Other potential sources include transportation as aggregates or by the finest grains 390 adhering to coarser particles during transport. Regardless of cause, these particles are unlikely to 391 yield meaningful information about wind regime variability or links to climate systems since they 392 do not yield a clear independent end member peak. 393 5.3 Aeolian dust dynamics in eastern Central Asia: links to atmospheric systems 394 Variations in grain size through time at NLK were largely driven by changes in wind strength, 395 without substantial influence of post-depositional pedogenesis. At NLK, grain size is therefore an 396 indicator of loess response to climatic systems. 397 The three end members are interpreted to represent different depositional processes which 398 operated throughout the accumulation of the deposit. The finer EM3 is interpreted to represent 399 constant background dust, which continued to accumulate throughout periods of relative stability 400 and pedogenesis. The coarser populations, EM1 and EM2, were transported by low-level winds 401 during major dust storms. EM1 is most likely the most sensitive recorder of wind intensity, since 402 EM2 is less sensitive to wind speeds than EM1 by observation of variations in EM2 proportions 403 throughout L1S1 and L1L2 (Fig. 4). 404 405 Fig. 7 Comparison between EM1 grain size variability and the timing of glacial advances in the Tien 406 Shan (Koppes et al. 2008; Owen and Dortch, 2014); stable oxygen isotope variations from the 407 Greenland ice cores (Rasmussen et al., 2014); mean grain size (MGS) record of the Jingyuan loess 408 section from the CLP (Sun et al., 2010) and; U-ratio (15.6‒63.4 µm/5.61‒15.6 µm) of the SE 409 Kazakhstan loess (Machalett et al., 2008). 5-point running average was performed for the intervals 410 with higher sedimentary rate on EM1 curve (red line). 411 412 From published OSL data (Song et al., 2015), we used linear regression (Stevens et al., 2016) 413 to construct age–depth relationships over intervals of visually similar sedimentation rate (Fig. S4 414 and Table S1). We assessed the degree of correlation between wind strength variability in the Ili 415 Valley (NLK), as represented by the proportions of EM1, with the stable oxygen isotope record from 416 the Greenland ice cores representing North Atlantic paleoclimate (Rasmussen et al., 2014), the mean 417 grain size (MGS) record of the Jingyuan loess section from the CLP (Sun et al., 2010), U-ratio 418 (15.6‒63.4 µm/5.61‒15.6 µm) of the Remizovka loess section in SE Kazakhstan (Machalett et al.,

419 2008), and glacial advances in the Tian Shan (Owen and Dortch, 2014;Koppes et al., 2008) (Fig. 7). 420 EM1 occurs in higher proportions during mid-MIS3, with a higher rate of sedimentary 421 accumulation (Fig. 7). Glaciers in the region expanded during early- and late-MIS3 (Owen and 422 Dortch, 2014). The apparent chronological link between increased primary loess accumulation and 423 glacial expansion in the region contrasts with trends elsewhere indicating increased dust 424 accumulation during dry-windy glacial conditions, and pedogenesis under comparatively wetter 425 interglacial conditions (Stevens et al., 2013;Sun et al., 2010;Ding et al., 2002;Dodonov and 426 Baiguzina, 1995). Our observations suggest a seesaw relationship between increased loess 427 accumulation and glacial expansion during MIS3 (Fig. 7), a model supported by Youn et al. (2014). 428 Moisture availability appears to be the dominant factor controlling glacier growth in Central Asia, 429 especially for glaciers in the Tian Shan (Zech, 2012;Koppes et al., 2008). We infer, therefore, that 430 moisture had an important impact on accumulation of dust in the study area during MIS3 in 431 particular. 432 Central Asia is variably influenced by the Asian monsoon from the south (Dettman et al., 433 2001;Cheng et al., 2012), the mid-latitude westerlies (Vandenberghe et al., 2006), the Siberian high- 434 pressure systems from the northeast (Youn et al., 2014), and the polar front from the north 435 (Machalett et al., 2008). The Asian high mountains largely inhibit the intrusion of Asian (Indian and 436 East Asian) monsoons to the region, since the Ili Valley is sheltered to the northeast, east and south. 437 Studies of the oxygen isotopic composition of precipitation in the Tian Shan Mountains region 438 support this geographic situation by indicating a stronger connection with westerly circulation than 439 with the Asian summer monsoon (Liu et al., 2015;Chen et al., 2016). 440 Modern satellite data indicates that dust storm development in Ili river valley is closely linked 441 with southward-moving high-latitude air masses (Ye et al., 2003). The large, cold Siberian High 442 pressure system is at the north-northeast of our study area, centring between 40°N and 65°N, 80°E 443 and 120°E (cf. Fig. 3 in Huang et al. (2011)). The Siberian anticyclone dominates winter and spring 444 climate over Eurasia (Sahsamanoglou et al., 1991;Savelieva et al., 2000;Panagiotopoulos et al., 445 2005;Gong and Ho, 2002;Obreht et al., 2017). Although the influence of the Siberian High has been 446 shown to decrease westward from the CLP (Vandenberghe et al., 2006), wind strength and frequency 447 over the Aral Sea in western central Asia during the Holocene was nevertheless associated with the 448 intensity of the Siberian High pressure system (Feng et al., 2011;Sorrel et al., 2007). Obreht et al. 449 (2017) even hypothesized increased influence of the Siberian High during MIS 3 over the Lower 450 Danube Basin in SE Europe, although this has yet to be substantiated. Moreover, the Siberian High 451 was considered to be one of the most important influences on dust deposition based on the results 452 of long-term monitoring over Central Asia between 2003 and 2010 (Groll et al., 2013). 453 Increases in modal grain-size from the CLP are also linked to a strengthened East Asian winter 454 monsoon due to intensification of the Siberian High (Ding et al., 1995;Hao et al., 2012). Therefore, 455 the grain-size record from the Chinese loess is a likely indicator of Siberian High intensity. We use 456 the Jingyuan loess section as a point of comparison in our study, because it is a high resolution 457 record located in the northwestern CLP, with high sedimentation rate, and thus the likelihood of 458 preservation of millennial-scale oscillations. We compared secular trends between the EM1 459 proportions and MGS data from Jingyuan over the last glacial period (Sun et al., 2010). Similarities 460 can be observed (Fig. 7); coarser grain sizes and higher sedimentation rates are observed during 461 mid-MIS3 (Sun et al., 2010), with the opposite occurring in early- and late-MIS3. This supports a 462 common Eurasian atmospheric forcing pattern - the Siberian High - driving the climate evolution of

463 the Ili Basin and CLP during that time period. 464 By comparison, the Last Glacial Maximum (LGM) witnesses significantly different trends, 465 despite increased sedimentation rates (Sun et al., 2010) (Fig. 7). EM1 proportions decrease 466 particularly during the early LGM. We attribute this to a reduction in sediment supply, possibly 467 linked to permafrost development in the Ili Basin and Kazakhstan steppe (Fig. 1) (Zhao et al., 468 2014;Vandenberghe et al., 2014). Reduced sediment supply therefore limits the degree to which 469 grain-size characteristics can reliably indicate wind strength during the LGM. 470 Machalett et al. (2008), presenting data from the Remizovka site in the more open western Ili 471 Basin, argued that the Arctic polar front, expanding southward in winter and retracting northward 472 in summer, most likely increased the frequency and strength of cyclonic storms due to higher 473 temperature and humidity gradients created between colder polar air and warmer tropical air 474 (Harman, 1991). They hypothesized that this climate system was the predominant influence on dust 475 transport and loess accumulation during cold phases along the Kyrgyz (southern) Tian Shan 476 piedmont. While this may have been the case at Remizovka, it is unlikely to have affected NLK in 477 the eastern Ili Basin, however, since the eastern basin is much more sheltered due to the position of 478 the mountain ranges (Fig. 1a). 479 To assess spatial variability in climatic influence across the Ili Basin, we compare EM1 curve 480 with U-ratio (15.6‒63.4 µm/5.61‒15.6 µm) of the polar-front-influenced Remizovka loess. We 481 observe minimal similarities in the curves. These disparities suggest that two different atmospheric 482 forcing patterns controlled loess accumulation from one end of the Ili Basin to the other. The 483 differences appear to be particularly clear over MIS3 (Fig. 7), although problems with chronological 484 integrity at the Remizovka site need to be resolved (Fitzsimmons et al., 2016) before we can argue 485 this with confidence. In addition, U-ratios decrease during the LGM (Fig. 7), supporting our 486 hypothesis that the development of permafrost limits the availability of source sediments for loess 487 in this region. 488 We argue that the Siberian high-pressure system exerts a significant influence on wind 489 dynamics and loess deposition in the eastern Ili Basin. It is evident that the strongest winds at NLK 490 mainly blow from the west (Table S2), although northerly high-latitude air masses with potential 491 for short-term dust transport can enter the Ili Basin by deflection around the northern Tianshan 492 mountains (Fig. S5). 493 Enhanced evaporation, coupled with strengthened westerly winds, would bring more humid 494 and warmer conditions to ACA during the Holocene (Zhang et al., 2016). Karger et al. (2016) 495 reconstructed the dynamics of the westerlies in the Ili Basin, proposing a rain belt which seasonally 496 migrates towards the south and north in autumn and summer, respectively. A strengthened Siberian 497 High would push the mid-latitude Westerlies pathways further to the south, resulting in comparably 498 drier conditions in northeastern Central Asia (e.g. Tian Shan) but wetter conditions in southwestern 499 Central Asia (Pamir) (Lei et al., 2014;Wolff et al., 2017). The intensity and geographical position of 500 the Siberian High would most likely impact precipitation and atmospheric circulation patterns 501 (meridional or zonal) in the mid-latitudes of Central Asia (Panagiotopoulos et al., 2005). It is 502 therefore most likely that the mid-latitude Westerlies controlled broad-scale patterns of moisture 503 variation across ACA broadly (Huang et al., 2015;Li et al., 2011;Cai et al., 2017), whereas the 504 eastern Ili Basin experienced the combined influence of the Siberian High and mid-latitude 505 Westerlies system. 506 Comparison of EM1 proportions with variability in GISP δ18O suggests that our grain-size

507 proxy data may correlate with abrupt events, such as North Atlantic Heinrich events H1 to H6 (Fig. 508 7), although this correlation cannot yet be better constrained due to limitations in the chronological 509 dataset. Some of the peaks in EM1 curve correspond to troughs in GISP δ18O curve (black arrows 510 in Fig. 7) outside Heinrich events, yet many do not (pink dashed lines in Fig. 7). Potential causes of 511 this discrepancy may lie in variability in local source availability and wind dynamics at certain 512 points in time. 513 Comparisons between the eastern Ili Basin and Chinese Loess Plateau loess further elucidates 514 complexity in the climatic signal preserved in the ACA. The NLK EM1 proportions in the Ili Basin 515 yield lower variability than the Jingyuan MGS on the CLP, particularly during H2 and H5 (Fig. 7). 516 We attribute these differences to local source sediment availability at NLK. EM1 supply to NLK 517 was reduced during H2 due to the development of permafrost, and during H5 due to increased 518 vegetation cover associated with more humid conditions inhibiting coarse-grain entrainment (Fig. 519 7). By contrast, the relatively more arid mid-MIS 3, indicated by glacial retreat in the Tian Shan, 520 may have decreased vegetation cover and increased entrainment potential and transport to NLK (Fig. 521 7); these conditions and this trend was also observed in the NE-TP (Vriend et al., 2011). The 522 differences may be because the loess records in our study area represent a response not only to 523 hemispheric climate systems, but also to local influences such as local atmospheric circulation and 524 topography. Since the sedimentary response to changing climate conditions in more arid Central 525 Asia is different to that of the more temperate European loess (Rousseau et al., 2007), we must be 526 careful about investigating the mechanisms of aeolian dynamics and loess accumulation in our 527 paleoclimatic interpretations of ACA loess archives. 528 Many studies have speculate that millennial-scale oscillations represent a teleconnection 529 between the North Atlantic and East Asia (e.g. Porter and An, 1995; Yang and Ding, 2014), although 530 the dynamics involved are poorly understood. Porter and An (1995) and Sun et al. (2012) suggested, 531 based on CLP loess physical characteristics, that a strong influence from the westerlies resulted in 532 transport of the North Atlantic signal to East Asia. Conversely, Yang and Ding (2014) proposed that 533 millennial-scale North Atlantic climate signals might have been transmitted to the Siberian High via 534 the Barents and Kara Sea ice sheets, and were propagated eastwards to the CLP via the winter 535 monsoon system. In the western CLP (Chen et al., 1997a), for example, evidence of millennial-scale 536 (likely Heinrich) events are preserved within the loess stratigraphy during phases of strong winter 537 monsoon in China; however, not all of the strong winter monsoon events in China correlate with 538 Heinrich events in the North Atlantic, so challenging the Yang and Ding (2014) hypothesis. 539 Stronger datasets from Central Asia may provide the missing link for understanding climate 540 teleconnections between the two extreme ends of the Eurasian continent. In doing so, however, the 541 scale of the “Central Asian” region must be taken into account. At Darai Kalon in Tajikistan, 1200 542 km southwest of NLK, the mid-latitude westerlies clearly have a strong influence on dust transport 543 and loess accumulation; Atlantic signals are clearly identified in grain size variations, especially 544 during full glacial phases (Vandenberghe et al., 2006). Since the CLP lies at a similar latitude to 545 Darai Kalon, mid-latitude Westerlies have the potential to transport North Atlantic climate signals 546 to East Asia. By contrast, since NLK is located substantially further north than Darai Kalon and the 547 CLP, the Siberian High exerts a greater influence on wind dynamics and therefore loess deposits. A 548 strengthened Siberian High would effect a southward shift of the mid-latitude Westerlies pathways; 549 under such conditions, NLK would be less strongly influenced by the mid-latitude westerlies. This 550 argument is further supposed by the lack of correlation between NLK EM1 proportions and GISP

551 δ18O values during relatively mild interstadial periods (Dansgaard-Oeschger cycles) when the mid- 552 latitude westerlies shift northwards (Fig. 7). Therefore, NLK provides a strategic location for 553 investigating the potential role of the Siberian High in transmitting North Atlantic climate signals 554 to East Asia. The preservation of North Atlantic several millennial-scale Heinrich events at NLK 555 supports the argument for the influence of the Siberian High as argued by Yang and Ding (2014). 556 Conclusion 557 Our data from NLK in the eastern Ili Basin provide a paleoenvironmental record over the last 558 c. 70 ky. The magnetic properties of the loess do not correlate with pedogenesis in this section; 559 rather, wind strength is mainly responsible for variations in physical characteristics over the last 560 glacial period. 561 Three grain-size end members were identified at NLK: EM1 (mode size at 47.5 µm), EM2 562 (33.6 µm) and EM3 (18.9 µm). They each indicate different kinds of depositional processes which 563 operated throughout the accumulation of the loess. EM1 and EM2 represent grain-size fractions 564 transported from proximal sources in short-term, near-surface suspension during dust outbreaks, 565 and may have the same origin. While wind strength controls their relative proportions, EM1 is the 566 most sensitive recorder of wind strength. EM3 represents continuous background dust under non- 567 dust storm conditions. 568 The Siberian High-pressure system predominates in the eastern Ili Basin during cold phases, 569 which leads to dust transport and increased loess accumulation at NLK. Many rapid cooling events, 570 including 6 Heinrich events, were imprinted in the NLK loess. Our grain size data support the 571 argument that the Siberian High plays a significant role in transporting North Atlantic climatic 572 signals to East Asia via ice sheets in the high northern latitudes. 573 Acknowledgements 574 The project is supported by the National Basic Research Program of China (Nos: 575 2016YFA0601902, 2013CB955904), Natural Science Foundation of China (Nos: 41572162, 576 41290253), and International Partnership Program of Chinese Academy of Science [No: 577 132B61KYS20160002]. The authors thank Yun Li and Junchao Dong from Institute of Earth 578 Environment, Chinese Academy of Sciences, for their assistances in sampling and experiment, and 579 Jia Li from Xiamen University for her assistance in mathematical treatment. 580 References 581 Antoine, P., Rousseau, D.-D., Moine, O., Kunesch, S., Hatté, C., Lang, A., Tissoux, H., and Zöller, 582 L.: Rapid and cyclic aeolian deposition during the Last Glacial in European loess: a high-resolution 583 record from Nussloch, Germany, Quaternary Sci Rev, 28, 2955-2973, 2009. 584 Beget, J. E., and Hawkins, D. B.: Influence of Orbital Parameters on Pleistocene Loess Deposition 585 in Central Alaska, Nature, 337, 151-153, DOI 10.1038/337151a0, 1989. 586 Bokhorst, M. P., Vandenberghe, J., Sumegi, P., Lanczont, M., Gerasimenko, N. P., Matviishina, Z. 587 N., Markovic, S. B., and Frechen, M.: Atmospheric circulation patterns in central and eastern 588 Europe during the Weichselian Pleniglacial inferred from loess grain-size records, Quatern Int, 234, 589 62-74, 10.1016/j.quaint2010.07.018, 2011. 590 Bond, G., Heinrich, H., Broecker, W., Labeyrie, L., Mcmanus, J., Andrews, J., Huon, S., Jantschik, 591 R., Clasen, S., Simet, C., Tedesco, K., Klas, M., Bonani, G., and Ivy, S.: Evidence for massive 592 discharges of icebergs into the North Atlantic Ocean during the last glacial period, Nature, 360, 593 245–249, 1992. 594 Bronger, A., and Heinkele, T.: Mineralogical and clay mineralogical aspects of loess research,

595 Quatern Int, 7, 37-51, 1990. 596 Buggle, B., Hambach, U., Glaser, B., Gerasimenko, N., Markovic, S., Glaser, I., and Zöller, L.: 597 Stratigraphy, and spatial and temporal paleocliamtic trends in Southeastern/Eastern European loess- 598 paleosol sequences, Quatern Int, 196, 86–106, 2009. 599 Buggle, B., Hambach, U., Muller, K., Zoller, L., Markovic, S. B., and Glaser, B.: Iron mineralogical 600 proxies and Quaternary climate change in SE-European loess-paleosol sequences, Catena, 117, 4- 601 22, 10.1016/j.catena.2013.06.012, 2014. 602 Cai, Y. J., Chiang, J. C. H., Breitenbach, S. F. M., Tan, L. C., Cheng, H., Edwards, R. L., and An, Z. 603 S.: Holocene moisture changes in western China, Central Asia, inferred from stalagmites, 604 Quaternary Sci Rev, 158, 15-28, 2017. 605 Chen, F., Bloemendal, J., Wang, J., Li, J., and Oldfield, F.: High-resolution multi-proxy climate 606 records from Chinese loess: evidence for rapid climatic changes over the last 75 kyr, 607 Palaeogeography, Palaeoclimatology, Palaeoecology, 130, 323-335, 1997a. 608 Chen, F. H., Bloemendal, J., Wang, J. M., Li, J. J., and Oldfield, F.: High-resolution multi-proxy 609 climate records from Chinese loess, evidence for rapid climatic changes over the last 75 kyr, 610 Palaeogeogr Palaeocl, 130, 323-335, 1997b. 611 Chen, F. H., Jia, J., Chen, J. H., Li, G. Q., Zhang, X. J., Xie, H. C., Xia, D. S., Huang, W., and An, 612 C. B.: A persistent Holocene wetting trend in arid central Asia, with wettest conditions in the late 613 Holocene, revealed by multi-proxy analyses of loess-paleosol sequences in Xinjiang, China, 614 Quaternary Sci Rev, 146, 134-146, 2016. 615 Cheng, H., Zhang, P. Z., Spotl, C., Edwards, R. L., Cai, Y. J., Zhang, D. Z., Sang, W. C., Tan, M., 616 and An, Z. S.: The climatic cyclicity in semiarid-arid central Asia over the past 500,000 years, 617 Geophys Res Lett, 39, Artn L01705 618 10.1029/2011gl050202, 2012. 619 Dansgaard, W., Johnsen, S. J., Clausen, H. B., Hvidberg, C. S., and Steffensen, J. P.: Evidence for 620 general instability of past climate from a 250-kyr ice-core record, Nature, 364, 218– 220, 1993. 621 Derbyshire, E., Meng, X. M., and Kemp, R. A.: Provenance, transport and characteristics of modern 622 aeolian dust in western Gansu Province, China, and interpretation of the Quaternary loess record, 623 Journal of Arid Environments, 39, 497-516, DOI 10.1006/jare.1997.0369, 1998. 624 Dettman, D. L., Kohn, M. J., Quade, J., Ryerson, F. J., Ojha, T. P., and Hamidullah, S.: Seasonal 625 stable isotope evidence for a strong Asian monsoon throughout the past 10.7 m.y., Geology, 29, 31- 626 34, Doi 10.1130/0091-7613(2001)029<0031:Ssiefa>2.0.Co;2, 2001. 627 Ding, Z., Liu, T., Rutter, N. W., Yu, Z., Guo, Z., and Zhu, R.: Ice-volume forcing of East Asian 628 winter monsoon variations in the past 800,000 years, Quaternary Res, 44, 149-159, 1995. 629 Ding, Z. L., Sun, J. M., Rutter, N. W., Rokosh, D., and Liu, T. S.: Changes in sand content of loess 630 deposits along a north-south transect of the Chinese Loess Plateau and the implications for desert 631 variations, Quaternary Research, 52, 56-62, DOI 10.1006/qres.1999.2045, 1999. 632 Ding, Z. L., Ranov, V., Yang, S. L., Finaev, A., Han, J. M., and Wang, G. A.: The loess record in 633 southern Tajikistan and correlation with Chinese loess, Earth Planet Sc Lett, 200, 387-400, doi: DOI: 634 10.1016/S0012-821X(02)00637-4, 2002. 635 Ding, Z. L., Derbyshire, E., Yang, S. L., Sun, J. M., and Liu, T. S.: Stepwise expansion of desert 636 environment across northern China in the past 3.5 Ma and implications for monsoon evolution, 637 Earth Planet Sc Lett, 237, 45-55, 10.1016/j.epsl.2005.06.036, 2005. 638 DiPietro, L. M., Driese, S. G., Nelson, T. W., and Harvill, J. L.: Variations in late Quaternary wind

639 intensity from grain-size partitioning of loess deposits in the Nenana River Valley, Alaska, 640 Quaternary Research, 87, 258-274, 2017. 641 Dodonov, A. E., and Baiguzina, L. L.: Loess stratigraphy of Central Asia: Palaeoclimatic and 642 palaeoenvironmental aspects, Quaternary Sci Rev, 14, 707-720, http://dx.doi.org/10.1016/0277- 643 3791(95)00054-2, 1995. 644 Dodonov, A. E., Sadchikova, T. A., Sedov, S. N., Simakova, A. N., and Zhou, L. P.: 645 Multidisciplinary approach for paleoenvironmental reconstruction in loess-paleosol studies of the 646 Darai Kalon section, Southern Tajikistan, Quatern Int, 152, 48-58, 10.1016/j.quaint.2005.12.001, 647 2006. 648 Feng, Z. D., Ran, M., Yang, Q. L., Zhai, X. W., Wang, W., Zhang, X. S., and Huang, C. Q.: 649 Stratigraphies and chronologies of late Quaternary loess-paleosol sequences in the core area of the 650 central Asian arid zone, Quatern Int, 240, 156-166, doi: 10.1016/j.quaint.2010.10.019, 2011. 651 Fitzsimmons, K. E., Sprafke, T., Zielhofer, C., Günter, C., Deom, J. M., Sala, R., and Iovita, R.: 652 Loess accumulation in the Tian Shan piedmont: Implications for palaeoenvironmental change in 653 arid Central Asia, Quatern Int, http://dx.doi.org/10.1016/j.quaint.2016.1007.1041, 2016. 654 Folk, R. L., and Ward, W. C.: Brazos River bar: a study in the significance of grain size parameters, 655 Journal of Sedimentary Research, 27, 3-26, 1957. 656 Gong, D. Y., and Ho, C. H.: The Siberian High and climate change over middle to high latitude Asia, 657 Theor Appl Climatol, 72, 1-9, 2002. 658 Grimley, D. A., and Arruda, N. K.: Observations of magnetite dissolution in poorly drained soils, 659 Soil Sci, 172, 968-982, 10.1097/ss.0b013e3181586b77, 2007. 660 Groll, M., Opp, C., and Aslanov, I.: Spatial and temporal distribution of the dust deposition in 661 Central Asia - results from a long term monitoring program, Aeolian Res, 9, 49-62, 2013. 662 Hao, Q., Wang, L., Oldfield, F., Peng, S., Qin, L., Song, Y., Xu, B., Qiao, Y., Bloemendal, J., and 663 Guo, Z.: Delayed build-up of Arctic ice sheets during 400,000-year minima in insolation variability, 664 Nature, 490, 393-396, 2012. 665 Hao, Q. Z., Oldfield, F., Bloemendal, J., and Guo, Z. T.: Particle size separation and evidence for 666 pedogenesis in samples from the Chinese Loess Plateau spanning the past 22 m.y., Geology, 36, 667 727-730, 10.1130/G24940a.1, 2008. 668 Harman, J. R.: Synoptic Climatology of the Westerlies: Process and Patterns, Association of 669 American Geographers, Washington, DC, 1991. 670 Heller, F., and Liu, T. S.: Magnetism of Chinese Loess Deposits, Geophys J Roy Astr S, 77, 125-&, 671 DOI 10.1111/j.1365-246X.1984.tb01928.x, 1984. 672 Heller, F., Shen, C. D., Beer, J., Liu, X. M., Liu, T. S., Bronger, A., Suter, M., and Bonani, G.: 673 Quantitative Estimates of Pedogenic Ferromagnetic Mineral Formation in Chinese Loess and 674 Paleoclimatic Implications, Earth Planet Sc Lett, 114, 385-390, Doi 10.1016/0012-821x(93)90038- 675 B, 1993. 676 Heller, F., and Evans, M. E.: Loess magnetism, Reviews of Geophysics, 33, 211–240, 1995. 677 Huang, W., Chen, J. H., Zhang, X. J., Feng, S., and Chen, F. H.: Definition of the core zone of the 678 "westerlies-dominated climatic regime", and its controlling factors during the instrumental period, 679 Sci China Earth Sci, 58, 676-684, 10.1007/s11430-015-5057-y, 2015. 680 Huang, X. T., Oberhansli, H., von Suchodoletz, H., and Sorrel, P.: Dust deposition in the Aral Sea: 681 implications for changes in atmospheric circulation in central Asia during the past 2000 years, 682 Quaternary Sci Rev, 30, 3661-3674, 2011.

683 Jacobs, P. M., Mason, J. A., and Hanson, P. R.: Mississippi Valley regional source of loess on the 684 southern Green Bay Lobe land surface, Wisconsin, Quaternary Research, 75, 574-583, 685 10.1016/j.yqres.2011.02.003, 2011. 686 Jia, J., Xia, D. S., Wei, H. T., Wang, B., and Liu, X. B.: A magnetic investigation of a loess/paleosol 687 sequences record in Ili area, Front Earth Sci-Prc, 4, 259-268, 10.1007/s11707-010-0115-4, 2010. 688 Jia, J., Xia, D. S., Wang, B., Wei, H. T., and Liu, X. B.: Magnetic investigation of Late Quaternary 689 loess deposition, Ili area, China, Quatern Int, 250, 84-92, 10.1016/j.quaint.2011.06.018, 2012. 690 Karger, D. N., Conrad, O., Böhner, J., Kawohl, T., Kreft, H., Soria-Auza, R. W., Zimmermann, N., 691 Linder, H. P., and Kessler, M.: Climatologies at high resolution for the Earth land surface areas, 692 arXiv:1607.00217, 2016. 693 Konert, M., and Vandenberghe, J.: Comparison of laser grain size analysis with pipette and sieve 694 analysis: A solution for the underestimation of the clay fraction, Sedimentology, 44, 523-535, DOI 695 10.1046/j.1365-3091.1997.d01-38.x, 1997. 696 Koppes, M., Gillespie, A. R., Burke, R. M., Thompson, S. C., and Stone, J.: Late Quaternary 697 glaciation in the Kyrgyz Tien Shan, Quaternary Sci Rev, 27, 846-866, 698 10.1016/j.quascirev.2008.01.009, 2008. 699 Lei, Y. B., Tian, L. D., Bird, B. W., Hou, J. Z., Ding, L., Oimahmadov, I., and Gadoev, M.: A 2540- 700 year record of moisture variations derived from lacustrine sediment (Sasikul Lake) on the Pamir 701 Plateau, Holocene, 24, 761-770, 2014. 702 Li, G. Q., Rao, Z. G., Duan, Y. W., Xia, D. S., Wang, L. B., Madsen, D. B., Jia, J., Wei, H. T., Qiang, 703 M. R., Chen, J. H., and Chen, F. H.: Paleoenvironmental changes recorded in a luminescence dated 704 loess/paleosol sequence from the Tianshan Mountains, arid central Asia, since the Penultimate 705 Glaciation, Earth Planet Sc Lett, 448, 1-12, 10.1016/j.epsl.2016.05.008, 2016a. 706 Li, J.: Climate in Xinjiang, in, China Meteorological Press, Beijing, 1-205 (in Chinese), 1991. 707 Li, X., Zhao, K., Dodson, J., and Zhou, X.: Moisture dynamics in central Asia for the last 15 kyr: 708 new evidence from Yili Valley, Xinjiang, NW China, Quaternary Sci Rev, 30, 3457-3466, 2011. 709 Li, Y., Song, Y., Lai, Z., Han, L., and An, Z.: Rapid and cyclic dust accumulation during MIS 2 in 710 Central Asia inferred from loess OSL dating and grain-size analysis, Scientific Reports, 6, DOI: 711 10.1038/srep32365, 2016b. 712 Li, Y., Song, Y. G., Chen, X. L., Li, J. C., Mamadjanov, Y., and Aminov, J.: Geochemical 713 composition of Tajikistan loess and its provenance implications, Palaeogeogr Palaeocl, 446, 186- 714 194, 10.1016/j.palaeo.2016.01.025, 2016c. 715 Lin, Y. C., Mu, G. J., Xu, L. S., and Zhao, X.: The origin of bimodal grain-size distribution for 716 aeolian deposits, Aeolian Research, 20, 80-88, 10.1016/j.aeolia.2015.12.001, 2016. 717 Liu, Q. S., Deng, C. L., Torrent, J., and Zhu, R. X.: Review of recent developments in mineral 718 magnetism of the Chinese loess, Quaternary Sci Rev, 26, 368-385, 10.1016/j.quascirev.2006.08.004, 719 2007. 720 Liu, Q. S., Roberts, A. P., Larrasoana, J. C., Banerjee, S. K., Guyodo, Y., Tauxe, L., and Oldfield, 721 F.: Environmental Magnetism: Principles and Applications, Reviews of Geophysics, 50, Artn 722 Rg4002 723 10.1029/2012rg000393, 2012. 724 Liu, T. S.: Loess and the environment, Ocean Press, Beijing, 1985b. 725 Liu, X., Liu, Z., Lü, B., Marković, S., Chen, J., Guo, H., Ma, M., Zhao, G., and Feng, H.: The 726 magnetic properties of Serbian loess and its environmental significance, Chinese Science Bulletin,

727 1-11, 2013. 728 Liu, X. K., Rao, Z. G., Zhang, X. J., Huang, W., Chen, J. H., and Chen, F. H.: Variations in the 729 oxygen isotopic composition of precipitation in the Tianshan Mountains region and their 730 significance for the Westerly circulation, J Geogr Sci, 25, 801-816, 2015. 731 Liu, X. M., Shaw, J., Liu, T. S., Heller, F., and Yuan, B. Y.: Magnetic Mineralogy of Chinese Loess 732 and Its Significance, Geophys J Int, 108, 301-308, DOI 10.1111/j.1365-246X.1992.tb00859.x, 1992. 733 Liu, X. M., Rolph, T., Bloemendal, J., Shaw, J., and Liu, T. S.: Remanence characteristics of 734 different magnetic grain size categories at Xifeng, central Chinese Loess Plateau, Quaternary 735 Research, 42, 162-165, 1994. 736 Liu, X. M., Hesse, P., Rolph, T., and Begét, J. E.: Properties of magnetic mineralogy of Alaskan 737 loess: evidence for pedogenesis, Quatern Int, 62, 93-102, 1999. 738 Lu, H. Y., and An, Z. S.: Pretreatment methods in loess–palaeosol granulometry, Chinese Science 739 Bulletin, 42, 237-240, 1997. 740 Lu, H. Y., and An, Z. S.: Paleoclimatic significance of grain size of loess-palaeosol deposit in 741 Chinese Loess Plateau, Sci China Ser D, 41, 626-631, Doi 10.1007/Bf02878745, 1998. 742 Machalett, B., Frechen, M., Hambach, U., Oches, E. A., Zoller, L., and Markovic, S. B.: The loess 743 sequence from Remisowka (northern boundary of the Tien Shan Mountains, Kazakhstan) - Part I: 744 Luminescence dating, Quatern Int, 152, 192-201, 10.1016/j.quaint.2005.12.014, 2006. 745 Machalett, B., Oches, E. A., Frechen, M., Zoller, L., Hambach, U., Mavlyanova, N. G., Markovic, 746 S. B., and Endlicher, W.: Aeolian dust dynamics in central Asia during the Pleistocene: Driven by 747 the long-term migration, seasonality, and permanency of the Asiatic polar front, Geochem Geophy 748 Geosy, 9, Artn Q08q09 749 10.1029/2007gc001938, 2008. 750 Maher, B., and Thompson, R.: Paleoclimatic significance of the mineral magnetic record of the 751 Chinese loess and paleosols, Quaternary Research, 37, 155–170, 1992. 752 Maher, B. A., and Taylor, R. M.: Formation of Ultrafine-Grained Magnetite in Soils, Nature, 336, 753 368-370, DOI 10.1038/336368a0, 1988. 754 Maher, B. A., and Thompson, R.: Paleorainfall reconstructions from pedogenic magnetic 755 susceptibility variations in the Chinese loess and paleosols, Quaternary Research, 44, 383-391, DOI 756 10.1006/qres.1995.1083, 1995. 757 Maher, B. A.: Magnetic properties of modern soils and Quaternary loessic paleosols: paleoclimatic 758 implications, Palaeogeogr Palaeocl, 137, 25-54, Doi 10.1016/S0031-0182(97)00103-X, 1998. 759 Maher, B. A.: Palaeoclimatic records of the loess/palaeosol sequences of the Chinese 760 Loess Plateau, Quaternary Sci Rev, 154, 23-84, 2016. 761 Mason, J. A., Jacobs, P. M., Greene, R. S. B., and Nettleton, W. D.: Sedimentary aggregates in the 762 Peoria Loess of Nebraska, USA, Catena, 53, 377-397, 10.1016/S0341-8162(03)00073-0, 2003. 763 Muhs, D. R., and Bettis, E. A.: Quaternary loess-paleosol sequences as examples of climate-driven 764 sedimentary extremes, Special Papers-Geological Society of America, 53-74, 2003. 765 Muhs, D. R.: The geologic records of dust in the Quaternary, Aeolian Research, 9, 3-48, 2013. 766 Nawrocki, J., Wojcik, A., and Bogucki, A.: The magnetic susceptibility record in the Polish and 767 western Ukrainian loess-palaeosol sequences conditioned by palaeoclimate, Boreas, 25, 161-169, 768 1996. 769 Nottebaum, V., Lehmkuhl, F., Stauch, G., Hartmann, K., Wunnemann, B., Schimpf, S., and Lu, H. 770 Y.: Regional grain size variations in aeolian sediments along the transition between Tibetan

771 highlands and north-western Chinese deserts - the influence of geomorphological settings on aeolian 772 transport pathways, Earth Surf Proc Land, 39, 1960-1978, 10.1002/esp.3590, 2014. 773 Nottebaum, V., Stauch, G., Hartmann, K., Zhang, J. R., and Lehmkuhl, F.: Unmixed loess grain size 774 populations along the northern Qilian Shan (China): Relationships between geomorphologic, 775 sedimentologic and climatic controls, Quatern Int, 372, 151-166, 10.1016/j.quaint.2014.12.071, 776 2015. 777 Obreht, I., Zeeden, C., Schulte, P., Hambach, U., Eckmeier, E., Timar-Gabor, A., and Lehmkuhl, F.: 778 Aeolian dynamics at the Orlovat loess-paleosol sequence, northern Serbia, based on detailed textural 779 and geochemical evidence, Aeolian Research, 18, 69-81, 2015. 780 Obreht, I., Hambach, U., Veres, D., Zeeden, C., Bösken, J., Stevens, T., Marković, S. B., Klasen, N., 781 Brill, D., and Burow, C.: Shift of large-scale atmospheric systems over Europe during late MIS 3 782 and implications for Modern Human dispersal, Scientific Reports, 7, 2017. 783 Owen, L. A., and Dortch, J. M.: Nature and timing of Quaternary glaciation in the Himalayan- 784 Tibetan orogen, Quaternary Sci Rev, 88, 14-54, 10.1016/j.quascirev.2013.11.016, 2014. 785 Panagiotopoulos, F., Shahgedanova, M., Hannachi, A., and Stephenson, D. B.: Observed trends and 786 teleconnections of the Siberian high: A recently declining center of action, J Climate, 18, 1411-1422, 787 2005. 788 Porter, S. C., and An, Z. S.: Correlation between climate events in the North-Atlantic and China 789 during Last Glaciation, Nature, 375, 305-308, Doi 10.1038/375305a0, 1995. 790 Prins, M. A., and Vriend, M.: Glacial and interglacial eolian dust dispersal patterns across the 791 Chinese Loess Plateau inferred from decomposed loess grain-size records, Geochem Geophy Geosy, 792 8, Artn Q07q05 793 10.1029/2006gc001563, 2007. 794 Prins, M. A., Vriend, M., Nugteren, G., Vandenberghe, J., Lu, H. Y., Zheng, H. B., and Weltje, G. J.: 795 Late Quaternary aeolian dust input variability on the Chinese Loess Plateau: inferences from 796 unmixing of loess grain-size records, Quaternary Sci Rev, 26, 230-242, 797 10.1016/j.quascirev.2006.07.002, 2007. 798 Prins, M. A., Zheng, H. B., Beets, K., Troelstra, S., Bacon, P., Kamerling, I., Wester, W., Konert, 799 M., Huang, X. T., Ke, W., and Vandenberghe, J.: Dust supply from river floodplains: the case of the 800 lower Huang He (Yellow River) recorded in a loess-palaeosol sequence from the Mangshan Plateau, 801 J Quaternary Sci, 24, 75-84, 10.1002/jqs.1167, 2009. 802 Pye, K.: Aeolian Dust and Dust Deposits, in, Academic Press, London, 29-62, 1987. 803 Pye, K., and Zhou, L. P.: Late Pleistocene and Holocene Aeolian Dust Deposition in North China 804 and the Northwest Pacific-Ocean, Palaeogeogr Palaeocl, 73, 11-23, Doi 10.1016/0031- 805 0182(89)90041-2, 1989. 806 Pye, K.: The nature, origin and accumulation of loess, Quaternary Sci Rev, 14, 653-667, Doi 807 10.1016/0277-3791(95)00047-X, 1995. 808 Qiang, M., Lang, L., and Wang, Z.: Do fine-grained components of loess indicate westerlies: 809 Insights from observations of dust storm deposits at Lenghu (Qaidam Basin, China), Journal of Arid 810 Environments, 74, 1232-1239, 10.1016/j.jaridenv.2010.06.002, 2010. 811 Qin, X. G., Cai, B. G., and Liu, T. S.: Loess record of the aerodynamic environment in the east Asia 812 monsoon area since 60,000 years before present, J Geophys Res-Sol Ea, 110, Artn B01204 813 10.1029/2004jb003131, 2005. 814 Rasmussen, S. O., Bigler, M., Blockley, S. P., Blunier, T., Buchardt, S. L., Clausen, H. B., Cvijanovic,

815 I., Dahl-Jensen, D., Johnsen, S. J., Fischer, H., Gkinis, V., Guillevic, M., Hoek, W. Z., Lowe, J. J., 816 Pedro, J. B., Popp, T., Seierstad, I. K., Steffensen, J. P., Svensson, A. M., Vallelonga, P., Vinther, B. 817 M., Walker, M. J. C., Wheatley, J. J., and Winstrup, M.: A stratigraphic framework for abrupt 818 climatic changes during the Last Glacial period based on three synchronized Greenland ice-core 819 records: refining and extending the INTIMATE event stratigraphy, Quaternary Sci Rev, 106, 14-28, 820 10.1016/j.quascirev.2014.09.007, 2014. 821 Rea, D. K., Leinen, M., and Janecek, T. R.: Geologic Approach to the Long-Term History of 822 Atmospheric Circulation, Science, 227, 721-725, DOI 10.1126/science.227.4688.721, 1985. 823 Rea, D. K., and Hovan, S. A.: Grain-Size Distribution and Depositional Processes of the Mineral 824 Component of Abyssal Sediments - Lessons from the North Pacific, Paleoceanography, 10, 251- 825 258, Doi 10.1029/94pa03355, 1995. 826 Rousseau, D. D., Sima, A., Antoine, P., Hatté, C., Lang, A., and Zöller, L.: Link between European 827 and North Atlantic abrupt climate changes over the last glaciation, Geophys Res Lett, 34, 22, 2007. 828 Sahsamanoglou, H. S., Makrogiannis, T. J., and Kallimopoulos, P. P.: Some Aspects of the Basic 829 Characteristics of the Siberian Anticyclone, Int J Climatol, 11, 827-839, 1991. 830 Savelieva, N. I., Semiletov, I. P., Vasilevskaya, L. N., and Pugach, S. P.: A climate shift in seasonal 831 values of meteorological and hydrological parameters for Northeastern Asia, Prog Oceanogr, 47, 832 279-297, 2000. 833 Schettler, G., Shabunin, A., Kemnitz, H., Knoeller, K., Imashev, S., Rybin, A., and Wetzel, H. U.: 834 Seasonal and diurnal variations in dust characteristics on the northern slopes of the Tien Shan - 835 Grain-size, mineralogy, chemical signatures and isotope composition of attached nitrate, Journal of 836 Asian Earth Sciences, 88, 257-276, 2014. 837 Smalley, I. J., Mavlyanova, N. G., Rakhmatullaev, K. L., Shermatov, M. S., Machalett, B., Dhand, 838 K. O., and Jefferson, I. F.: The formation of loess deposits in the Tashkent region and parts of Central 839 Asia; and problems with irrigation, hydrocollapse and soil erosion, Quatern Int, 152, 59-69, 840 10.1016/j.quaint.2005.12.002, 2006. 841 Song, Y., Zeng, M., Chen, X., Li, Y., Chang, H., An, Z., and Guo, X.: Abrupt climatic events 842 recorded by the Ili loess during the last glaciation in Central Asia: Evidence from grain-size and 843 minerals, Journal of Asian Earth Sciences, in press, https://doi.org/10.1016/j.jseaes.2017.1010.1040, 844 2017a. 845 Song, Y. G., Shi, Z. T., Fang, X. M., Nie, J. S., Naoto, I., Qiang, X. K., and Wang, X. L.: Loess 846 magnetic properties in the Ili Basin and their correlation with the Chinese Loess Plateau, Sci China 847 Earth Sci, 53, 419-431, 10.1007/s11430-010-0011-5, 2010. 848 Song, Y. G., Li, C. X., Zhao, J. D., Cheng, P., and Zeng, M. X.: A combined luminescence and 849 radiocarbon dating study of the Ili loess, Central Asia, Quat Geochronol, 10, 2-7, 850 10.1016/j.quageo.2012.04.005, 2012. 851 Song, Y. G., Chen, X. L., Qian, L. B., Li, C. X., Li, Y., Li, X. X., Chang, H., and An, Z. S.: 852 Distribution and composition of loess sediments in the Ili Basin, Central Asia, Quatern Int, 334, 61- 853 73, 10.1016/j.quaint.2013.12.053, 2014. 854 Song, Y. G., Lai, Z. P., Li, Y., Chen, T., and Wang, Y. X.: Comparison between luminescence and 855 radiocarbon dating of late Quaternary loess from the Ili Basin in Central Asia, Quat Geochronol, 30, 856 405-410, 10.1016/j.quageo.2015.01.012, 2015. 857 Song, Y. G., Li, Y., Li, Y., An, Z., Cheng, L., Sun, H., and Rustam, O.: North Atlantic Abrupt Climate 858 Signals during the Last Glacial Period in Central Asia: Evidences from Aeolian Loess Sediments,

859 Acta Geologica Sinica - English Edition, 91, 1135-1136, 2017b. 860 Sorrel, P., Oberhansli, H., Boroffka, N., Nourgaliev, D., Dulski, P., and Rohl, U.: Control of wind 861 strength and frequency in the Aral Sea basin during the late Holocene, Quaternary Res, 67, 371-382, 862 2007. 863 Stevens, T., Adamiec, G., Bird, A. F., and Lu, H. Y.: An abrupt shift in dust source on the Chinese 864 Loess Plateau revealed through high sampling resolution OSL dating, Quaternary Sci Rev, 82, 121- 865 132, 10.1016/j.quascirev.2013.10.014, 2013. 866 Stevens, T., Buylaert, J. P., Lu, H., Thiel, C., Murray, A., Frechen, M., Yi, S., and Zeng, L.: Mass 867 accumulation rate and monsoon records from Xifeng, Chinese Loess Plateau, based on a 868 luminescence age model, J Quaternary Sci, 31, 391-405, 2016. 869 Sun, D. H., Bloemendal, J., Rea, D. K., Vandenberghe, J., Jiang, F. C., An, Z. S., and Su, R. X.: 870 Grain-size distribution function of polymodal sediments in hydraulic and aeolian environments, and 871 numerical partitioning of the sedimentary components, Sediment Geol, 152, 263-277, Pii S0037- 872 0738(02)00082-9 873 Doi 10.1016/S0037-0738(02)00082-9, 2002. 874 Sun, D. H., Chen, F. H., Bloemendal, J., and Su, R. X.: Seasonal variability of modern dust over the 875 Loess Plateau of China, J Geophys Res-Atmos, 108, Artn 4665 876 10.1029/2003jd003382, 2003. 877 Sun, D. H., Bloemendal, J., Rea, D. K., An, Z. S., Vandenberghe, J., Lu, H. Y., Su, R. X., and Liu, 878 T. S.: Bimodal grain-size distribution of Chinese loess, and its palaeoclimatic implications, Catena, 879 55, 325-340, 10.1016/S0341-8162(03)00109-7, 2004. 880 Sun, D. H.: Supper-fine grain size components in Chinese loess and their palaeoclimatic implication, 881 Quaternary Sciences, 26, 928-936 (in Chinese with English abstract), 2006. 882 Sun, Y., Lu, H., and An, Z.: Grain size distribution of quartz isolated from Chinese loess1 paleosol, 883 Chinese Science Bulletin, 45, 2296-2298, 2000a. 884 Sun, Y., Lu, H. Y., and An, Z. S.: Grain size of loess, palaeosol and Red Clay deposits on the Chinese 885 Loess Plateau: Significance for understanding pedogenic alteration and palaeomonsoon evolution, 886 Palaeogeogr Palaeocl, 241, 129-138, 10.1016/j.palaeo.2006.06.018, 2006. 887 Sun, Y., Clemens, S. C., Morrill, C., Lin, X., Wang, X., and An, Z.: Influence of Atlantic meridional 888 overturning circulation on the East Asian winter monsoon, Nature Geoscience, 5, 46-49, 2012. 889 Sun, Y. B., Lu, H. Y., and An, Z. S.: Grain size distribution of quartz isolated from Chinese 890 loess/paleosol, Chinese Science Bulletin, 45, 2296-2298, Doi 10.1007/Bf02886372, 2000b. 891 Sun, Y. B., Wang, X. L., Liu, Q. S., and Clemens, S. C.: Impacts of post-depositional processes on 892 rapid monsoon signals recorded by the last glacial loess deposits of northern China, Earth Planet Sc 893 Lett, 289, 171-179, 10.1016/j.epsl.2009.10.038, 2010. 894 Terhorst, B., Ottner, F., and Wriessnig, K.: Weathering intensity and pedostratigraphy of the Middle 895 to Upper Pleistocene loess/palaeosol sequence of Wels-Aschet (Upper Austria), Quatern Int, 265, 896 142-154, 2012. 897 Tsoar, H., and Pye, K.: Dust Transport and the Question of Desert Loess Formation, Sedimentology, 898 34, 139-153, DOI 10.1111/j.1365-3091.1987.tb00566.x, 1987. 899 Ujvari, G., Kok, J. F., Varga, G., and Kovacs, J.: The physics of wind-blown loess: Implications for 900 grain size proxy interpretations in Quaternary paleoclimate studies, Earth-Sci Rev, 154, 247-278, 901 10.1016/j.earscirev.2016.01.006, 2016. 902 Vandenberghe, J., Renssen, H., van Huissteden, K., Nugteren, G., Konert, M., Lu, H. Y., Dodonov,

903 A., and Buylaert, J. P.: Penetration of Atlantic westerly winds into Central and East Asia, Quaternary 904 Sci Rev, 25, 2380-2389, 10.1016/j.quascirev.2006.02.017, 2006. 905 Vandenberghe, J.: Grain size of fine-grained windblown sediment: A powerful proxy for process 906 identification, Earth-Sci Rev, 121, 18-30, 10.1016/j.earscirev.2013.03.001, 2013. 907 Vandenberghe, J., French, H. M., Gorbunov, A., Marchenko, S., Velichko, A. A., Jin, H., Cui, Z., 908 Zhang, T., and Wan, X.: The Last Permafrost Maximum (LPM) map of the Northern Hemisphere: 909 permafrost extent and mean annual air temperatures, 25–17 ka BP, Boreas, 43, 652-666, 2014. 910 Varga, G.: Similarities among the Plio-Pleistocene terrestrial aeolian dust deposits in the World and 911 in Hungary, Quatern Int, 234, 98-108, 10.1016/j.quaint.2010.09.011, 2011. 912 Vriend, M., and Prins, M. A.: Calibration of modelled mixing patterns in loess grain-size 913 distributions: an example from the north-eastern margin of the Tibetan Plateau, China, 914 Sedimentology, 52, 1361-1374, 10.1111/j.1365-3091.2005.00743.x, 2005. 915 Vriend, M.: Lost in loess: Late Quaternary eolian dust dispersal patterns across Central China 916 inferred from decomposed loess grain-size records, Amsterdam, VU University, The Netherlands, 917 53 pp., 2007. 918 Vriend, M., Prins, M. A., Buylaert, J. P., Vandenberghe, J., and Lu, H. Y.: Contrasting dust supply 919 patterns across the north-western Chinese Loess Plateau during the last glacial-interglacial cycle, 920 Quatern Int, 240, 167-180, 10.1016/j.quaint.2010.11.009, 2011. 921 Wang, H., Mason, J. A., and Balsam, W. L.: The importance of both geological and pedological 922 processes in control of grain size and sedimentation rates in Peoria Loess, Geoderma, 136, 388-400, 923 10.1016/j.geoderma.2006.04.005, 2006. 924 Weltje, G. J.: Endmember modeling of compositional data: numerical–statistical algorithms for 925 solving the explicit mixing problem, Journal of Mathematical Geology, 29, 503–549, 1997. 926 Weltje, G. J., and Prins, M. A.: Genetically meaningful decomposition of grain-size distributions, 927 Sediment Geol, 202, 409-424, 2007. 928 Wolff, C., Plessen, B., Dudashvilli, A. S., Breitenbach, S. F., Cheng, H., Edwards, L. R., and Strecker, 929 M. R.: Precipitation evolution of Central Asia during the last 5000 years, Holocene, 27, 142-154, 930 2017. 931 Xiao, J., Porter, S. C., An, Z. S., Kumai, H., and Yoshikawa, S.: Grain-Size of Quartz as an Indicator 932 of Winter Monsoon Strength on the Loess Plateau of Central China during the Last 130,000-Yr, 933 Quaternary Research, 43, 22-29, DOI 10.1006/qres.1995.1003, 1995. 934 Yang, F., Zhang, G. L., Yang, F., and Yang, R. M.: Pedogenetic interpretations of particle-size 935 distribution curves for an alpine environment, Geoderma, 282, 9-15, 2016. 936 Yang, S. L., Ding, F., and Ding, Z. L.: Pleistocene chemical weathering history of Asian arid and 937 semi-arid regions recorded in loess deposits of China and Tajikistan, Geochim Cosmochim Ac, 70, 938 1695-1709, 10.1016/j.gca.2005.12.012, 2006. 939 Yang, S. L., and Ding, Z. L.: Advance-retreat history of the East-Asian summer monsoon rainfall 940 belt over northern China during the last two glacial-interglacial cycles, Earth Planet Sc Lett, 274, 941 499-510, 10.1016/j.epsl.2008.08.001, 2008. 942 Yang, S. L., and Ding, Z. L.: A 249 kyr stack of eight loess grain size records from northern China 943 documenting millennial-scale climate variability, Geochem Geophy Geosy, 15, 798-814, 944 10.1002/2013gc005113, 2014. 945 Ye, W.: Characteristics of physical environment and conditions of loess formation in Yili area, 946 Xinjiang, Arid Land Geography, 22, 9-16 (in Chinese), 1999.

947 Ye, W., Sang, C., and Zhao, X.: Spatial-temporal distribution of loess and source of dust in Xinjiang, 948 Journal of Desert Research, 23, 514-520 (in Chinese), 2003. 949 Youn, J. H., Seong, Y. B., Choi, J. H., Abdrakhmatov, K., and Ormukov, C.: Loess deposits in the 950 northern Kyrgyz Tien Shan: Implications for the paleoclimate reconstruction during the Late 951 Quaternary, Catena, 117, 81-93, 10.1016/j.catena.2013.09.007, 2014. 952 Yu, S. Y., Colman, S. M., and Li, L. X.: BEMMA: A Hierarchical Bayesian End-Member Modeling 953 Analysis of Sediment Grain-Size Distributions, Math Geosci, 48, 723-741, 10.1007/s11004-015- 954 9611-0, 2016. 955 Zech, R.: A late Pleistocene glacial chronology from the Kitschi-Kurumdu Valley, Tien Shan 956 (Kyrgyzstan), based on Be-10 surface exposure dating, Quaternary Research, 77, 281-288, 957 10.1016/j.yqres.2011.11.008, 2012. 958 Zeeden, C., Hambach, U., Veres, D., Fitzsimmons, K., Obreht, I., Bösken, J., and Lehmkuhl, F.: 959 Millennial scale climate oscillations recorded in the Lower Danube loess over the last glacial period, 960 Palaeogeography, Palaeoclimatology, Palaeoecology, (in press), 2016. 961 Zhang, X. J., Jin, L. Y., Huang, W., and Chen, F. H.: Forcing mechanisms of orbital-scale changes 962 in winter rainfall over northwestern China during the Holocene, Holocene, 26, 549-555, 963 10.1177/0959683615612569, 2016. 964 Zhang, X. Y., Arimoto, R., and An, Z. S.: Glacial and interglacial patterns for Asian dust transport, 965 Quaternary Sci Rev, 18, 811-819, 1999. 966 Zhao, L., Jin, H., Li, C., Cui, Z., Chang, X., Marchenko, S. S., Vandenberghe, J., Zhang, T., Luo, 967 D., and Guo, D.: The extent of permafrost in China during the local Last Glacial Maximum (LLGM), 968 Boreas, 43, 688-698, 2014. 969 Zhou, L. P., Oldfield, F., Wintle, A. G., Robinson, S. G., and Wang, J. T.: Partly Pedogenic Origin 970 of Magnetic Variations in Chinese Loess, Nature, 346, 737-739, DOI 10.1038/346737a0, 1990. 971

Fig01

Fig02

Fig03

Fig04

Fig05

Fig06

Fig07